Some Quantile Regression Models for Zero-Inflated Continuous Data with Applications

Date of Award

7-10-2022

Publication Type

Dissertation

Degree Name

Ph.D.

Department

Mathematics and Statistics

Keywords

Continuous data, Generalized exponential, Kumaraswamy, Non-parametric model, Quantile regression, Two-part model

Supervisor

A.Hussein

Supervisor

S.Hossain

Rights

info:eu-repo/semantics/embargoedAccess

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Abstract

This thesis develops new models for quantile regression for zero-inflated, non-negative data defined on intervals of the form [0, a) where a = 1 or ∞. The first host of models proposed in this thesis rely on two parametric distributions known as Generalized exponential and Kumaraswamy distributions, respectively, for data defined in [0, ∞) and [0, 1). The second set of models are semi-parametric models in which the zeros are modeled through a logistic regression model and the positive part of the data are modeled by using the usual linear/non-linear quantile regression models. We perform intensive Monte Carlo simulations to assess the performance of all proposed methods and we illustrate the utility of the methodologies by applying them to data sets on vehicle corrosion.

Share

COinS